Skip to content

[Bugfix] [ROCm] [UX]: revert Flex attention backend #29371

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
DarkLight1337 merged 1 commit into from
Nov 25, 2025
Merged

[Bugfix] [ROCm] [UX]: revert Flex attention backend #29371

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions tests/v1/attention/test_rocm_attention_backends_selection.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,12 @@ def mock_on_gfx9():
@pytest.mark.parametrize(
"env_vars, selected_backend, expected_backend_path",
[
# Test Case: Explicit FLEX_ATTENTION backend
(
{},
"FLEX_ATTENTION",
AttentionBackendEnum.FLEX_ATTENTION.get_path(),
),
# Test Case 1: Default (no env vars, no explicit backend)
(
{},
Expand Down
4 changes: 4 additions & 0 deletions vllm/platforms/rocm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -262,6 +262,10 @@ def get_attn_backend_cls(
f"is not MLA type while requested for MLA backend."
)

if selected_backend == AttentionBackendEnum.FLEX_ATTENTION:
logger.info("Using FlexAttention backend.")
return AttentionBackendEnum.FLEX_ATTENTION.get_path()

if selected_backend == AttentionBackendEnum.TRITON_ATTN:
logger.info("Using Triton Attention backend on V1 engine.")
return AttentionBackendEnum.TRITON_ATTN.get_path()
Expand Down
Loading